Myth #5: Economists, Using Charts or High Speed Computer Models, Can Accurately Forecast the Future

Recorded by the Mises Institute in the mid-1980s, The Mises Report provided radio commentary from leading non-interventionists, economists, and political scientists. In this program, we present another part of “Ten Great Economic Myths”. This material was prepared by Murray N. Rothbard.

The problem of forecasting interest rates illustrates the pitfalls of forecasting in general. People are contrary cusses whose behavior, thank goodness, cannot be forecast precisely in advance. Their values, ideas, expectations, and knowledge change all the time, and change in an unpredictable manner. What economist, for example, could have forecast (or did forecast) the Cabbage Patch Kid craze of the Christmas season of 1983? Every economic quantity, every price, purchase, or income figure is the embodiment of thousands, even millions, of unpredictable choices by individuals.

Many studies, formal and informal, have been made of the record of forecasting by economists, and it has been consistently abysmal. Forecasters often complain that they can do well enough as long as current trends continue; what they have difficulty in doing is catching changes in trend. But of course there is no trick in extrapolating current trends into the near future. You don’t need sophisticated computer models for that; you can do it better and far more cheaply by using a ruler. The real trick is precisely to forecast when and how trends will change, and forecasters have been notoriously bad at that. No economist forecast the depth of the 1981–82 depression, and none predicted the strength of the 1983 boom.

The next time you are swayed by the jargon or seeming expertise of the economic forecaster, ask yourself this question: If he can really predict the future so well, why is he wasting his time putting out newsletters or doing consulting when he himself could be making trillions of dollars in the stock and commodity markets?

For more episodes, visit Mises.org/MisesReport